Building Brands Online


Published on

1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Building Brands Online

  1. 1. Sponsored byBUILDING DS BRANONLINE Interactive Branding:Best Practices in a Direct Response-Driven Media
  2. 2. Here’s to more noise and heavier traffic.Chevy was looking for new ways to reach its target. So Yahoo! presented the carmaker with the opportunity to betheir first Digital AdVentures partner, o ering an innovative new ad unit as part of a testing and learning initiative.Tracking showed that search volume soared almost immediately, with some models increasing over 700%.That’s the power of SCIENCE + ART + SCALE. Find out more at©2010. Yahoo! All Rights Reserved.
  3. 3. Sponsored by TABLE OF CONTENTSINTRODUCTION 4 CHARTS- Branding vs. direct response CHART 1 Brand dollars vs. direct dollars, 4- What brands are spending on interactive now, and where Barclays Capital CHART 2 All forms of TV viewing are increasing, 5PLANNING 9 The Nielsen Co.- Best Practice: Get a new funnel CHART 3 Online ad-spend data, The Nielsen Co. 6- Inventory evaluation tools CHART 4 Retail, CPG and public services show 6- Best Practice: Evaluate clutter of all media partners greatest growth, The Nielsen Co.- Best Practice: Use reach-buying techniques for brand CHART 5 Top brands online, ranked by share of 7 launches or repositionings spending, The Nielsen Co.- Best Practice: Understand and use audience cumes CHART 6 Top sites ranked by impressions, 8 to flight campaigns Q2 2010, The Nielsen Co.- Best Practices in usage of ad networks and exchanges CHART 7 Search, purchases by type of website, 10 Online Publishers AssociationMEASUREMENT 15 CHART 8 Publisher ad-clutter summary, 10- Clicks: Brands embrace response June 2010, comScore- Best Practice: Branding impact studies CHART 9 Audience cumes, comScore 11- Best Practice: Measure branding impact CHART 10 Top audience segments purchased, 12- Best Practice: Find an optimal online conversion path June 2010, Audience Science- Best Practice: Measure offline-sales impact over time CHART 11 Social network vs. portal share of 12- Best Practice: Use current GRP tools–support global total minutes, comScore initiatives to create better ones CHART 12 Custom ads on Facebook improve 13- Best Practice: Pre- and post-buy studies: Get what effectiveness, The Nielsen Co. you paid for CHART 13 Reach of ad networks, 14- Best Practice: Test for optimal frequency comScore- Best Practice: Employ brand-protection tools for CHART 14 Worst malware found, July-August 2010 15 context safety ClickFacts- Best Practice: Was the ad even seen? Above-the-fold CHART 15 Metrics brand marketers want, Bain/IAB 17 tools CHART 16 Brand metric glossary, Dynamic Logic 17- Best Practice: Re-marketing consumers- Best Practice: Incorporate interactive into media-mix CHART 17 Brand awareness increases with multiple 18 models media use, Dynamic Logic CHART 18 Online ad impact over time, comScore 20CREATIVE 23 CHART 19 Sample media plan, Atlas/comScore 21- Best Practices: Dynamic Logic, Nielsen IAG CHART 20 Pre- and post-testing, comScore 22- Best Practice: Don’t optimize creative to the click CHART 21 Best practices in frequency, 24- New creative options Insight Express CHART 22 Poorly performing campaigns 24CONCLUSION 26 negatively impact branding, Dynamic LogicFinding the magicThis document, and information contained therein, is the MORE ON ADAGE.COMcopyrighted property of Crain Communications Inc. andAdvertising Age (© Copyright 2010) and is for your personal, This is one in a series of white papers published bynon-commercial use only. You may not reproduce, display on a Advertising Age. To see other Ad Age white papers andwebsite, distribute, sell or republish this document, or the to obtain additional copies of this white paper, go toinformation contained therein, without prior written consent ofAdvertising Age. Copyright 2010 by Crain Communications Inc. rights reserved. | October 11, 2010 | 3
  4. 4. BUILDING BRANDS ONLINEINTRODUCTIONCHART 1 Brand dollars vs. direct dollars 2009 U.S. MEASURED MEDIA SPEND 2009 U.S. ONLINE MEDIA SPEND TOTAL: $147 billion TOTAL: $24 billion $6 billion DIRECT DIRECT $55 billion $18 billion $91 billion BRAND BRAND Source: Brand net analysis based on Barclays Capital, Think Equity Partners LLC, and DMAAfter 15 years of interactive, is it time for best practices?I BY KATHRYN KOEGEL 2011, brand marketers expected that 70% of dollars would go to print and TV, while direct-response marketers would spend nearlyfifteen years ago, a couple of engineers hung out in a basement an equal amount on online as on print and TV. When one considersin Georgia and created something—perhaps not magical, but that all forms of digital media tend to reach younger audiences thanabsolutely irrevocable in terms of media: the ad server. It enabled print and TV in a more highly engaged manner, online has a lot ofads to be dispersed, displayed and tracked on the burgeoning plat- explaining to do. Millions of media dollars are being left on the table.form then being called the World Wide Web. Since there was no The reality is that marketers have a serious challenge buying and“there there” in terms of physical media—no tapes or print copies planning interactive media.It’s too complicated,ignores basic preceptsto check ad delivery—the ad server would show the advertiser how of marketing (including the significance of creative and the value ofmany times an ad was viewed, and where. Given that this new context), and has for too long set itself apart from other media.medium had an infinite array of placement opportunities and That complexity prompted the IAB to act last month.potential inventory, the aim was to simplify the whole process. “Measurement is one of the key obstacles to growing spend in inter- The engineers were Kevin O’Connor and Dwight Merriman, and active,” said Sherrill Mane, IAB senior VP-industry services.“Unlessthey founded DoubleClick—now a part of Google. They were engi- we create a smooth supply chain so that online can be bought and soldneers,and weren’t concerned with building and selling brands.Yet they as simply as TV, we will be held back...It’s a business-process problemfocused on solving a crucial problem: the complexity of interactive that the entire ecosystem of the business has to take control of.” Themedia. Little did they know they were in the process of blowing apart IAB, with the Association of National Advertisers and Americanall of media, laying waste to the revenue streams of “traditional” pub- Association of Advertising Agencies, issued an RFP to consultinglishers and engendering a world of numerical complexity that few in firms to create “a structure for change.” Meanwhile, the Marketingthe media world can even claim to understand. Accountability Standards Board, which includes members Coca-Cola Most importantly,brand advertisers have seemingly been left out Co.,Publicis,Starcom and the ANA,is striving “to create a set of met-of this data-driven, digital-media revolution. According to a rics generally recognized as meaningful and predictive.”A third analysis of data from Barclay Capital, Think Equity ect involves Google and the ARF issuing an RFP for a project to makePartners and the Direct Marketing Association, online accounts for reach and frequency metrics comparable between online and TV.30% of the $55 billion spent on direct marketing, yet accounts for While the industry looks for answers, this report will address pit-only 6% of the $91 billion spent on branding in the 2009 (see falls online faces as a branding medium and simplify, simplify, sim-chart 1). The Interactive Advertising Bureau commissioned Bain to plify. Tools, techniques and creative options can make interactivesurvey 700 marketers in April 2010. When asked to look forward to media sing for brand marketers—we just need to adopt them.4 | October 11, 2010 |
  5. 5. Sponsored byCHART 2 All forms of TV viewing are increasing, but watching “linear” TV remains the dominant mode. NUMBER OF USERS 2+, Q1 2010 DIFFERENCE IN MONTHLY REACH FROM Q1 2009 Watching TV in the home 286 million +0.6% Using a Mobile Phone 229 million +18.1% Using the Internet on a PC 191 million +17.3% Watching Video on Internet 135 million +2.6% Watching Timeshifted TV 96 million +5.3% Mobile Subscribers Watching Video on a Mobile Phone 20 million +51.2% 0 100 200 300 Source: The Nielsen Co.Branding vs. Direct Responsethe worlds of branding and direct response are the Venus ing points will result in sales of how many rolls of toilet paper.and Mars of the ad world—or are they? One is all squishy and How do they know this? Media-mix modeling and scanner data.emotional, filled with beautiful sentiment that goes straight to TV buys for package goods are as much a data-driven business asthe heart. The other gets right to the point: click, slam, bam, online is.thank you, ma’am—you’ve just lowered your car insurance orwhitened your teeth. TV USAGE AND EFFECTIVENESS ARE NOT DIMINISHING AS AUDIENCES GO ONLINE That’s a gross simplification of how branding and direct Nielsen’s Three Screen Report is an essential read for any mar-response work, and how they have been irrevocably changed in keter. The data show that while television usage is changing andthe digital age. Perianne Grignon, former CMO of Sears and now includes more variables, TV likely will always be the dominantCMO and chief strategy officer for [x+1], a demand-side plat- medium in terms of consumer usage, and is actually increasing inform, acknowledges the divide but sees it as a somewhat artificial all forms (see chart 2). TV viewers, however, are getting older:one: “I actually think that pure brand practitioners and perform- while the average person 2+ spends more than 35 hours per weekance marketers measure similar things. They just do it in differ- watching TV, people 65+ spend more than 48 hours. In terms ofent ways, in different speeds, and call the measures different effectiveness, Joel Rubinson, former VP-research of thethings. At the heart of it is a desire for a product, and whether Advertising Research Foundation, finds TV is working just asthat’s called conversion or intent, it’s all leading to the same place: well. Marketers, however, are rightly concerned about certaina purchase.” trends, especially increasing fragmentation of reach, clutter and ad Now, to clear up a few misconceptions about direct and brand- skipping. Yet, the prices don’t go down, and in this year’s Upfronting, and media channels: alone, over $9 billion dollars of advertising was purchased.TELEVISION DOES MORE THAN JUST BRANDING, AND BRANDING DOES LEAD TO SALES MOST INTERACTIVE MEDIA CAN HAVE A BRANDING IMPACT, BEYOND DIRECT RESPONSEYes, those ads make people hum and feel good, but they also sell To Rubinson, all interactive has some kind of a branding impact,product—though that product is not sold directly through the and that includes search. “In the new world of media, brands getTV, in most cases. Brand marketers have some 50 years of expe- created in more subtle ways,” he said. “Branding effects arerience buying TV, and they know precisely how many gross rat- everywhere.” | October 11, 2010 | 5
  6. 6. BUILDING BRANDS ONLINECHART 3 Online ad spend data: Search has grown fastest; Display, Rich Media & Video also show positive trends. IAB internet ad revenue share by major format, 2004 through 2009 50% 40 HELP WANTED Classifieds went from 18% of total online ad spend in 2004 to 10% in 2009, 30 affecting total ad spend 20 10 ’05 ’06 ’09 ’08 ’04 ’07 ’05 ’05 ’05 ’05 ’05 ’06 ’09 ’06 ’09 ’06 ’09 ’06 ’09 ’06 ’09 ’08 ’08 ’08 ’08 ’08 ’04 ’04 ’04 ’04 ’04 ’07 ’07 ’07 ’07 ’07 0 SEARCH DISPLAY BANNERS CLASSIFIEDS RICH MEDIA LEAD GENERATION SPONSORSHIPS Source: The Nielsen Co.BRANDING ONLINE CAN MEAN ASKING FOR A CLICK CHART 4 Retail, CPG and public services show greatest growthOnline-ad creative has the capacity for the click, and brand mar-keters are taking advantage of this.That may not be their complete INDUSTRY Q2 2010 SOI* YOY % CHANGE SOI*measure of performance, but if they can get a response, they will.Whole generations now are familiar with the concept of interact- FINANCIAL SERVICES 23.4% 4%ing with brands and ads, and marketers are taking advantage of WEB MEDIA 15.8% -10%that in ways that engage and often entertain. According toMichael Cassidy, CEO of Undertone, a brand-focused ad network, RETAIL GOODS & SERVICES 13.8% 25% the majority of brand creative he sees is “brand response.” In the case of CPG, it’s CONSUMER GOODS 10.3% 41% often a drive to a coupon or a game; for TELECOMMUNICATIONS 8.3% -35% entertainment, a drive to a microsite or trailer. PUBLIC SERVICES 7.1% 71% The whole notion of “branding” as someCLICK TO COUPONS sort of consumer/product connection/value ENTERTAINMENT 5.1% -13%Duncan Hines ad system dictated by marketers is not rele-prompts response AUTOMOTIVE 4.5% 6% vant in the age of ubiquitous media andconsumer control. Interactive media have blown up the very HEALTH 3.7% -2%notion of what a brand is, making it more consumer-involved anddynamic. Brands can now be direct sellers, content producers, TRAVEL 3.5% -31%bloggers, tweeters and even friends without having to rely on BUSINESS TO BUSINESS 1.7% -32%media to deliver those messages. Consumers can seek out thosebrands, connect with them through social networks, tweet about SOFTWARE 1.6% -4%them, and instantaneously let all their friends know what theythink about them or what they plan to buy. HARDWARE & ELECTRONICS 1.2% -23% The very fact that we still talk about new and old media, inter-active and traditional, goes to the heart of the problem, says Note: * Share of impressions; Source: The Nielsen Co.Rishad Tobaccowala, chief strategy and innovation officer at6 | October 11, 2010 |
  7. 7. Sponsored byCHART 5 Foods make top spenders list along with perennials like loan consolidation, auto and wireless. Top brands ranked on share of estimated spending during first half 2010, U.S. RANK BRAND PARENT COMPANY 1ST HALF 2010 YOY % CHANGE SHARE OF ESTIMATED SPEND IN SHARE OF SPEND IN SHARE OF SPEND 1 LOWERMYBILLS.COM INC. Experian Group Ltd. 2.66% 235% 2 WYETH PHARMACEUTICALS Pfizer Inc. 1.32% 4,011% 3 CHEVROLET General Motors Corp. 0.98% 226% 4 AT&T WIRELESS SERVICES INC. AT&T Corp. 0.82% -44% 5 HILLSHIRE FARM Sara Lee Corp. 0.75% N/A 6 CLASSES USA INC. Experian Group Ltd. 0.71% 95% 7 TOYOTA Toyota Motor Corp. 0.67% 3% 8 LENDINGTREE.COM InterActiveCorp 0.58% 285% 9 FORD Ford Motor Co. 0.47% -56% 10 ORVILLE REDENBACHER ConAgra Foods Inc. 0.44% 1,890,271% 11 CHEETOS PepsiCo Inc. 0.39% 2,193% 12 CENTRUM Pfizer Inc. 0.38% 23,581% 13 NISSAN Nissan Motor Co. Ltd. 0.38% 91% 14 PET HEALTH & NUTRITION Procter & Gamble Co. 0.34% 3,760% 15 WIRELINE COMMUNICATIONS AT&T Corp. 0.31% -25% Source: The Nielsen Co.VivaKi. “People in the U.S. are in a post-digital age,” he said. the fact that display, rich media and video account for a smaller“People don’t stop and say, ‘Now I’m going to use analog media; share than search, it’s interesting how much entrepreneurial activ-now I’m going to use digital.’ The future doesn’t fit into the con- ity is focused on image-based ads. An entirely new sector hastainers of the past.” developed over the past three years in the automation of image- What are those containers? Old notions of TV as a bucket for based online advertising: we now have Demand Side Platforms,branding dollars, online as the receptacle for direct. “As marketers, Supply Side Platforms, Ad Exchanges, Data Optimizers and cloud-we need to live this change,” said Tobaccowala. “People discover, computing companies focused on simplifying the process andtransact, express and share—they don’t consume advertising. In the increasing the effectiveness of interactive advertising. Google is socurrent times, people are marketing to themselves.” What’s the bullish on display, it took the unusual step last month of launchingsolution, as he sees it? “Plan across media, deliver engagement, an ad campaign, including a Times Square billboard, to promotemeasure across, never use the word ‘digital.’” the value of its display network. When the house that search built starts talking up the power of display, and a whole lot of venture capitalists jump in to support automation of interactive display, it’sWhat and where brands are clear the market sees untapped potential. Given that so many of these companies are data-driven, you might assume that interac-spending on interactive now tive is becoming even more direct response-driven. Are brands embracing this hyped-up advertising world? Nielsen’s AdRelevance impression numbers reflect only CPM buys (notwhen we talk about brand-related advertising online, we typi- those purchased on direct-response metrics like cost-per-acquisi-cally are referring to visual ads such as display, rich media and tion and cost-per-click), and while Q1 2010 was flat versus the pre-video rather than search. Search, the ultimate “bottom funnel” vious year, Q2 was up 7%.activity (someone knows what they want to buy and wants to Nielsen’s online ad-impression data by category shows thatknow where to buy it or how much it costs), has continued to grow heavy “brand categories” are indeed buying into online (see charteven during the recession, as other forms of interactive advertising 4). Financial services remains the top category, though the declinetook a nosedive. Online classifieds were hardest hit (see chart 3), of financial ads online lead to stories of the “death of the banner”as data from the IAB indicates—no surprise given that recruitment throughout 2008 and 2009 (see reports The State of Display I andand real estate are two of classifieds biggest categories. Display was II). The top growth categories are retail (up 25%), consumer goodsstable, while rich media and video showed slight increases. Given (up 41%) and public services (up 71%). Automotive is stable, up | October 11, 2010 | 7
  8. 8. BUILDING BRANDS ONLINECHART 6 Social networks are the big growth story, but Yahoo still carries nearly double the volume. Top sites ranked by impressions during 2Q 2010, U.S. 2Q10 IMPRESSIONS (IN BILLIONS) Yahoo 156.92 Facebook 80.75 MSN 57.65 MySpace 39.31 The Weather Channel 11.83 10.90 10.21 FOXNEWS.COM CNBC 8.80 6.72 +16% YOY % change for Yahoo YouTube 6.71 TOP GAINERS TOP LOSERS Juno 6.20 146% -46% Facebook 119% -37% MSNBC 5.86 MSN 70% IMDb -34% CNN 4.70 66% NeoPets -31% 4.36 Juno 40% MSNBC -30% MySpace 24% -23% Google 3.90 The Weather Channel 32% -10% Verizon Online 3.47 NeoPets 31% CNN -7% YouTube 30% Google N/A Pogo 2.86 MSNBC 30% IMDb 2.60 Road Runner 23% 2.50 Verizon Online 18% Pogo 16% Road Runner 2.45 15% NeoPets 2.10 Fox Sports on MSN 9% CNBC 9% FOX Sports on MSN 1.76 1.67 1.58 1.49 0 50 100 150 200 Source: The Nielsen Co.6%. Interestingly, the telecommunications sector, one of the heavi- embracing interactive as a marketing medium.est spenders during the recession (all those smartphones!), declined The rest of the list also has positive stories for branding (autoby 35%. Telco numbers as a percent of total have decreased due to brands, for example, make up seven of the top 25). Who are theother types of marketers coming back into interactive.It also reflects of yore? and Lending Tree are perenni-the fact that cell phones are employing a lot of direct-response als of the top 10. Where are the dollars going? Clearly, brand mar-strategies online as conversions are the primary goal. keters continue looking for reach, but there are new ways for them to The top 15 spenders list for online advertisers (see chart 5) is get that reach: namely, social networks. Among the top sites rankedalways an interesting mishmash. In the early ’00s, the list was by display-ad impressions,the biggest gainers have been social-mediadominated by the dot-coms. Coming out of the last recession sites (see chart 6). It should be noted that this Nielsen data is blind to(2002-’03), Fortune 500 companies had begun to make the list. In the cost of those impressions. A report from comScore in June 2010the second quarter of 2010, Hillshire Farms, Orville Redenbacher, showed that social media cost-per-thousand rates (CPMs) were soCheerios and Centrum appear. Clearly, brand marketers are low the overall average CPM came down considerably.8 | October 11, 2010 |
  9. 9. Sponsored byPLANNING“Brand media planning has changed so much. It used to be about has impact? Yaakov Kimelfeld, senior VP-digital research and ana-reach and relevancy; now it’s about reach and relevancy in a world of lytics director at MediaVest, believes that in today’s media world,consumer control that is becoming more social and more mobile the purchase path is more complex and ever-changing than Briggs’layered on with shopper-marketing practices.” four funnels could ever accommodate. The group he runs at-PERIANNE GRIGNON, FORMER CMO OF SEARS AND CURRENT VP-MEDIA STRATEGY MediaVest conducts weekly surveys of a panel of consumers, whichAND CHIEF MARKETING OFFICER AT [X+1] helps develop media-path analysis on a per-product basis. Media consumption is changing so rapidly and radically that this kind of“Online suffers because it’s the Swiss Army Knife of media. It’s got a ongoing check-in with consumers yields valuable insights,toothpick, a magnifying glass and a knife. It can work in so many places.” Kimelfeld said.- REX BRIGGS, CEO OF MARKETING EVOLUTION For an industry futurist like Rishad Tobaccowala, who advocates blowing up the funnel and changing how agencies operate (“ThereBriggs’ observation about the multifaceted nature of online brings are too many that just are production houses for :30s—they shouldinto focus another long-held marketing precept, the notion of a lin- just go away”), media planning needs to start with going back to theear purchase funnel and its use in the planning process. If we con- human being.“The single most important thing we need to keep insider how consumers made purchase deci-sions 40 years ago—when typically, theywatched one of three TV networks, read a ‘The single most important thing we need to keep in minddaily newspaper in print (one which supplied is that we are marketing to people, not consumers orcoupons in the Thursday food section and customers.’ - Rishad Tobaccowala, chief strategy andSunday inserts) and listened to local radio sta-tions on their way to work each morning—it innovation officer VivaKi ,was all pretty simple. Each medium had itsstrengths (TV for branding/reach, print for branding and informa- mind is that we are marketing to people, not consumers or cus-tion/product consideration, newspapers and radio for local and pro- tomers,” he explained.motions). The yellow pages came in at the bottom, as consumers In the today’s world, discovering the distinct, changing path tolooked up where to buy products. purchase, and isolating people who have displayed intent, is certain- Over the past 15 years, online has pummeled that funnel, leav- ly a best it hopelessly leaking and generally useless. Various forms ofinteractive can work at different levels of the funnel. The 79% of BEST PRACTICE: INVENTORY EVALUATION THAT ASSESSES QUALITY OF PLACEMENTSconsumers who use the internet (Pew, August 2010) rarely pur- “Agencies are desperate to find something that works, and right nowchase a “considered” product like a car without getting more infor- the only inventory truly worth anything is premium. There’s actually amation online first. A significant number of purchases in categories finite amount of quality inventory. Premium pricing will go up if you justlike travel and electronics have shifted to online. Younger and more shut off the billions of impressions on Facebook.”male-skewing audiences tend to become aware of entertainment —DOROTHY YOUNG, FOUNDER OF THE GLASS BOX AND FORMER COO OF OGILVYONE AND CEO OF SILVER CARROTand even personal-care products online. Search has replaced yellowpages—and mobile search has made it available on the go, while “The solution to crap inventory is not to buy crap.”consumers are already actively shopping. Promotionally sensitive —YAAKOV KIMMELFELD, SENIOR VP-DIGITAL RESEARCH AND ANALYTICS DIRECTOR, MEDIAVESTconsumers are not only clipping but also clicking their way tocoupons. So what’s a media planner to do, seeing that hard-and-fast Perhaps no job in the online world is more thankless than the rolerules about which media work best for awareness, consideration and of the online planner. Thousands of sites (25,000 in Nielsen’s data-purchase decisions no longer exist? bank alone), so little time, so much less money.As Dorothy Young, the former COO of OgilvyOne who has also consulted with inter-BEST PRACTICE: GET A NEW FUNNEL—AND START UNDERSTANDING MORE ABOUT PEOPLE active agencies, points out: “We’ve created a world of such com-AND THEIR PURCHASE PROCESS ON A CATEGORY AND EVEN PER-BRAND BASIS plexity that agencies can’t make any money buying it—90% ofRex Briggs understands the appeal of simple diagrams for explain- the work of a media department is manual processing.”ing concepts like how people come to make a purchase. To him, the What should planners be looking at in the way of content to givefunnel should be broken into four types based on the life cycle of full branding value to the ads? It’s not so different from the qualitythe product and goals like acquisition/trial, brand building, main- measures that are applied to print. Considerations should includetenance and conversion/retention. the quality of the placement, using parameters such as clutter, ad Do we need even more finite funnels showing where interactive size, position on page (including whether the ad appears above the | October 11, 2010 | 9
  10. 10. BUILDING BRANDS ONLINE CHART 7 Portal channels more likely to lead to search; Media sites INVENTORY-EVALUATION TOOLS impact purchases Some shortcuts can help evaluate online inventory for quality. 20% BRAND SEARCHES PURCHASES ONLINE PUBLISHERS ASSOCIATION SITES 15 Members have to fulfill various qualifications, including a professional staff that edits content. Still, the OPA does not employ 10 any sort of “Good Content” seal, and parent companies, not individual sites, are listed as members, so no complete list of OPA 2% 5 14% C 8% BC 16%C sites is available. 5% C 8% COMSCORE 100, 250 OR 500 0 OPA (A) Portal Channels (B) Social Media (C) comScore’s ad-planning tool allows users to limit searches to only BASE Site Ratings (N=6222: 4185 OPA, 1544 Portal Channels, 493 Social Media). Q: Which of the following have you top sites by reach and allows filtering by types of content such as ever done as the result of seeing an ad on [SITE]. Searched for more information about the product or service, made a purchase? Superscripted capital letters (ABC) indicate statistically significant differences at 90% confidence adult, social, user-generated, etc. between site categories. Source: Online Publishers Association URL WHITELISTS Adnetik keeps massive lists of sites carrying advertising, and CHART 8 News sites are most cluttered. excludes ones with reported placement and ad-load time comScore’s Publisher Ad Clutter Summary report problems. DISPLAY ADS PER TOTAL PAGES GOOGLE AD PLANNER News/Information This free ad-planning tool ( is accessible to anyone. Top 1,000 properties are ranked by unique users, which doesn’t automatically denote quality but is an easy Sports way to filter out properties too small for an advertiser’s needs. Conversational Media NIELSEN Portals As of March, @plan was integrated into the media-planning tool. Claritas Prizm Clusters are also now integrated, so planners can Community rank properties by reach within chosen groups. 0.0 0.5 1.0 1.5 2.0 Source: comScore, June 2010fold), time spent on page and repeat visitation patterns of the site. sites have different behaviors to non-premium.” HeWith so many sites, that amounts to a ton of work. also believes that a finite amount of quality inventory is available Young hopes to help, and is in the midst of building an invento- out there.ry-scoring tool to take some of the pain out of planning online. The OPA attempted to get at this elusive halo effect in its studyShe’s positioning her new company, dubbed The Glass Box, as an “A Sense of Place,” conducted by Harris and released in Juneantidote to all that much of the online space has become. 2010. Consumers related their positive associations with a media Does context truly matter? It’s been a battle that premium brand to the advertisers on the site, expecting a certain level ofpublishers have fought since internet upstarts began flooding the quality from those advertisers. Consumers displayed the most loy-ad world with inventory. While someone like David Payne, alty to “branded” sites, and while portals were most likely to drivefounder of ShortTail Media (a short-lived, brand-focused ad net- search activity, branded sites were most likely to drive built on inventory from Online Publishers Association Social media appears to be a different media beast altogether (seemembers), felt he was spending too much time trying to push the chart 7).value of context on agencies, the majority of those interviewed forthis report were strong proponents of context. Tobaccowala, for BEST PRACTICE: EVALUATE CLUTTER OF ALL MEDIA PARTNERSone, notes: “Brands like to hang out with brands.” Clutter has long been a challenge of online advertising. The more Even those most associated with the data/direct-response ads per page, the more they are ignored. (One exception to that areperspective are coming around. Nathan Woodman, managing sites that are used for comparative shopping, especially for auto.director-COO, Adnetik, an ad trading desk, noted based on its data See Nielsen’s “Measuring Online Advertising Clutter: A Newanalysis, “We are starting to see that people that visit premium Perspective for Media Planning.”) Despite research from Nielsen,10 | October 11, 2010 |
  11. 11. Sponsored byDynamic Logic and Insight Express indicating that fewer ads were 1 BIG AD,more effective, in these tough times, some publishers have 4 CARSincreased ad loads. Which category most has to get a handle on the Chevroletprogram? News (see chart 8). featured four models Brand marketers should always ask publishers what the ad in thisload is on their web pages, in addition to ascertaining placement Yahoo richon the page. comScore now has a report available in AdMetrix media unitcalled the Publisher Ad Clutter Summary report, which offersdata on Average Display Ads Per Total Pages,Average Display AdsPer Visit, and more.BEST PRACTICE: USE REACH-BUYING TECHNIQUES FOR BRAND LAUNCHESOR REPOSITIONING CHART 9 Understanding how a site cumes its audience can help inFor the past decade, various sites have tried to create splashy the flighting of campaignsunits and inventory packages to help brand marketers supportlaunches or achieve quick reach.Here are some successful techniques. 120% REACH OF UNIQUES I Homepage takeovers First launched on Yahoo nearly 10 years ago, homepagetakeovers of major sites like CNN, ESPN,The New York Times andThe Wall Street Journal are the rough equivalent of running a full-page, back-of-a-section ad in a major newspaper. All of these sites DAY 2, 40% Yahoo and Facebook cumeoffer the opportunity to surround content with sequenced banners 100 their audiences at similar ratesor roadblocks, or to offer exclusive use of homepages. I Google Display Network blasts These can also reach a large audience in a short time period (usu-ally 24-76 hours). With a signature property like YouTube, theseblasts are the equivalent of prime-time sponsorships of network TVprograms. Google reports that brand marketers like 80InterContinental Hotels & Resorts, Infiniti and H&R Block havetaken advantage of them.What kind of reach can they achieve? TheInfiniti March Madness Blast delivered 79 million ad impressionsin four days, for example.The YouTube homepage placement alonedelivered 40 million views per day. I Yahoo Log-In Page Units 60 CARS.COM Introduced this summer, this mega-rich-media ad unit (1,400pixels wide) is placed on the third most-highly trafficked page on ESPNYahoo. About 26 million unique users in the U.S. each day visit FACEBOOK.COMthe log-in page that connects Yahoo subscribers with their e-mailaccounts, stock pages and other services. Chevrolet ran a campaign NEW YORK TIMESin June 2010 with four different executions that had users click on 40 DIGITALa link placed within the background art that takes them to a page STYLE.COMwith more information on the Chevrolet model displayed. YAHOO SITESBEST PRACTICE: UNDERSTAND AND USE AUDIENCE CUMESTO FLIGHT CAMPAIGNS APPROPRIATELY 20Planners typically buy sites based on potential reach. There’s asimple fallacy to that logic: You would have to buy every impres-sion that goes to every user over a given month to reach thatnumber. What can you do to increase reach of the potential audi-ence so buys are flighted accordingly? Tactics like homepagetakeovers typically reach heaviest users of a site with high fre- 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31quency, not the full potential of the site’s reach. A quick glance atchart 9 shows that a one-day takeover of Yahoo’s homepage DAY IN THE MONTHwould reach just under 40% of the site’s users. In contrast, it takes Source: comScore, May 2010The New York Times Digital about seven days to reach 40%. | October 11, 2010 | 11
  12. 12. BUILDING BRANDS ONLINEBEST PRACTICE: USE AUDIENCE TARGETING TO EXTEND REACH BEYOND CONTEXT CHART 10 Top Audience Segments purchased: June 2010OR TO INCREASE THE EFFICIENCY OF A BUY SEASONALWhen Tacoda introduced “behavioral targeting” in 2001, it was RANK AUDIENCE SEGMENT OUTDOOR ENTHUSIASTS,seen as a way to juice response rates—as in clicks. Behavioral tar- 1 AUTO > AUTO SHOPPERS VACATION TRAVELERSgeting was largely the domain of direct-response advertisers—or 2 LIFESTYLE > OUTDOOR ENTHUSIASTbrand marketers who were driving online conversions and buying HOT EVERY MONTHon a DR basis. It’s now had its own rebranding and emerged as 3 LIFESTYLE > ECO-AWARE INDIVIDUALS AUTO (4), CELL PHONE“audience targeting.” 4 TECHNOLOGY > CELL PHONE ENTHUSIASTS ENTHUSIASTS While audience targeting is still used heavily by DR advertis- 5 LIFESTYLE > PROUD PARENTS CUSTOMers, Audience Science CEO Jeff Hirsch reports marketers are using 6 CUSTOM > 1 OF THE TOP 10it even when they are not looking to drive a specific transaction. Itrepresents 30% of his customers, an increase from 2009, he 7 AUTO > ASIAN IMPORT BUYERSreports. Hirsch sees the technique as a crucial bridge to TV buying, 8 AUTO > LUXURY AUTOMOBILE BUYERSwhere the aim is to deliver a message to a sizable audience. “It’s asclose as we are going to get to a GRP (gross rating point),” he said. 9 TRAVEL > VACATION TRAVELERS According to Amanda Richman, executive VP-managing direc- 10 AUTO > AUTO ENTHUSIASTStor, digital at MediaVest, audience targeting does play a role in Source: Audience Science Quarterly Report, June, 2010brand buys as it “extends reach beyond that implied by context.” John Montgomery, COO of GroupM, views it as an answer toan efficiency problem. “Our brand clients are paying too much reach aggregation. In a given month, one to three of the top slotswhen they buy context directly from publishers—soap doesn’t typically go to custom audience segments; there are a few broadneed context,” he said. “We need to make online a more efficient targets marketers would recognize from TV (women 25-54, menbuy.” For him, the shift toward audience buying makes sense 18-34). Automotive clearly is using the technique to find in-mar-beyond just cost. “It’s intent rather than segment-based market- ket buyers at more affordable prices than those on auto sites anding, and uses the unique targeting capabilities of online,” he said. in auto context. It appears in the top 10 categories every month, So who buys audience targeting and how finite does it get? An usually in more than one permutation. Other top targets areanalysis of top targets bought on the Audience Science platform “gadget heads” and “cell phone enthusiasts.” For telco, it’s a mat-over the past year shows that it is a mix of very specific targeting ter of reach plus intent. (See chart 10 for the top targets purchased(custom segments requested by advertisers), intent marketing and from Audience Science in June 2010.)CHART 11 Consumers are spending more time on social networks, less time on portals. 30% SHARE OF GLOBAL TOTAL MINUTES SPENT ONLINE FACEBOOK YAHOO YOUTUBE GOOGLE MSN 8 8 8 7 7 4 6 7 7 20 4 5 5 5 7 7 6 6 6 6 6 6 5 5 5 5 5 10 5 5 5 5 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 1 1 1 1 1 1 1 1 1 1 1 1 1 0 June July August September October November December January February March April May June 2009 2010 Audience: Persons age 15+. Worldwide, home/work/university locations. Source: comScore Media Metrix12 | October 11, 2010 |
  13. 13. Sponsored byBEST PRACTICES IN USE OF SOCIAL MEDIA: REACH OR CONVERSATION? CHART 12 Nielsen study shows that custom ads on Facebook designed toAnyone looking at the user-growth curve of Facebook must stand in elicit “brand friends” work better than traditional adsawe before this latest player in the reach market (see chart 11, P. 12). Over the past year, the site morphed from a utility for twen- DIFFERENCE BETWEEN CONTROL GROUP AND EXPOSEDtysomethings into a mass-market destination that reaches 42% of 20% HOMEPAGE AD EXPOSUREthe online population, according to Pew. There are a whole lot of HOMEPAGE ADpeople out there posting status updates, but is advertising on social SOCIAL ADVOCACYnetworks the best possible use of this type of communication? The OPA study on the significance of various types of contentfor advertising shows that ads on social networks are not as effec- 10tive at driving product sales or even brand searches as other typesof content. Perhaps the environment for social-media ads is akin toe-mail: so personal and engrossing that ads are not particularlynoticed. To be sure, more research on the topic is needed. 0 Ad Recall Awareness Purchase Intent So, if the inventory is not as great in value as content, what is itgood for? Nielsen, which has developed a product in conjunction Source: Nielsen BrandLiftwith Facebook, called Facebook Brand Lift, points out in its report“Advertising Effectiveness: Understanding the Value of a SocialMedia Impression” that ads on Facebook are better served to gen- specialty, strong content and solid agency relationships. Nathanerate brand conversations (see chart 12). The basic point is that Woodman, COO of Adnetik, a trading desk spinoff from Havas,the same creative used to push a message elsewhere may not work cites NBC Universal’s decision to create its own ad network as a pos-as well in a social environment. “Brand Advocacy” ads, or ads itive step. In doing so, the media company joined Time Inc., Forbes,designed to drive conversation about a brand, are a better way to go. Martha Stewart Omnimedia and others in pooling company inven- tory to provide reach of quality audiences.Another key survival tac-BEST PRACTICES IN USAGE OF AD NETWORKS AND EXCHANGES: tic? The ability to deliver rich-media units that cannot be bought onPROCEED WITH EYES WIDE OPEN exchanges and delivering them at scale.Networks play a vital role in the online ecosystem,and have,since theday DoubleClick built its ad server, to be able to track and distribute HOW TO EVALUATE AN AD NETWORK FOR BRANDING USEonline media—and then built an ad network to sell it. Ad networks Here are key questions one should use to evaluate ad networks.are also one of the most controversial parts of the ecosystem for a I Buying Practicesvery simple fact: many operate black boxes of inventory where there How does the network secure its ad inventory? Does it buy fromis little concern for issues such as quality, clutter and appropriateness other networks? Ad exchanges? Direct from publishers? Is there anyof content. Perhaps even deeper issues are the lack of exclusivity of exclusivity of inventory? Buyers should be specific about the kindsinventory and the “gross” margins they receive for services. Among of inventory they want. One media buyer of pharma ads makes net-those interviewed for this white paper, the estimates of inventory works sign a document specifying inventory will not be purchasedflowing through ad networks bought and sold on exchanges range on exchanges, then monitors for accuracy and demands makegoods.from 75% to 90%. In effect, the same inventory is being passed back I Quality controland forth until it is sold at the lowest possible price. Others take issue Does the network guarantee ads will run above the fold? That allwith networks that charge excessive fees without adding any value. international impressions will be excluded? That there is noSays Dorothy Young,former COO of Ogilvy One and CEO of Silver “parked” inventory or social inventory you choose to exclude? UpCarrot: “Networks get 30% to 50% gross margins on the media; to 40% of publisher site traffic comes from outside of the U.S., andagencies could never charge that. Should the clients accept it?” ad networks are used as a clearinghouse for this inventory. Unless Networks can be a powerful part of simplifying the process of international users are your target, specify U.S.-traffic only inbuying interactive ads and achieving the reach demanded by brand- insertion orders. Parked sites (those sites that show up when a useroriented buys. Chart 13 (P. 14) shows networks as ranked by mistypes a URL) is inventory that often flows through networks—comScore (comScore notes these rankings are somewhat problem- especially ones bought on a cost-per-acquisition basis.atic due to the issue of nonexclusivity of inventory). How is the content in the network assessed? A common practice But with more than 300 networks and little differentiation is to use top sites in comScore as a proxy for quality—though thatbetween them, not to mention the increasing automation of sales reflects reach more than quality. Due to contracts, most highly val-through exchanges and DSPs, how many networks can the market ued publishers blind the inventory. How do networks back up anysupport? Those interviewed expressed varying levels of skepticism claims to the quality of sites that make up their inventory?about the future of networks. John Montgomery of GroupM thinks One of the most dangerous challenges of inventory that comesthe space will contract (and only needs a maximum of 30 networks) through networks—especially ad exchanges which have entirelyas more inventory flows into DSPs and exchanges, which are just in automated the buying process and where no physical contact occurstheir infancy now. Those that will thrive will likely have a vertical with the purchaser of the inventory—is malware. Malware can do | October 11, 2010 | 13
  14. 14. BUILDING BRANDS ONLINE CHART 13 Ad network reach, June 2009 vs. June 2010 TOTAL UNIQUE VISITORS (000) Yahoo Network AOL Advertising Google Ad Network ValueClick Networks 24/7 Real Media BIG 3 Microsoft Media Network U.S. Yahoo, Google and AOL run the Specific Media top networks for reach Turn Inc. AdBrite FOX Audience Network Collective Network Tribal Fusion Traffic Marketplace AudienceScience JUNE 2009 JUNE 2010 interCLICK 0 50,000 100,000 150,000 200,000 Source: Comscoresomething as simple as use an ad to infect a computer with a virus ties? Which ad server does it use, and how proficient is the networkor, in the worst case, scrape identities like credit card data from a at using it? Most importantly, do trafficking people understand theuser’s computer. Chart 14 (P. 15) shows some incidents of malware goals of the campaign and use the tools to optimize accordingly—andthat hit various sites this summer, as captured by ClickFacts. not just default to the direct-response metrics they know? I Brand Protection I Data usage Will the network guarantee ads will not run on undesirable sites Does the network work with third-party resources to enrich dataand issue a makegood should that happen? Will it issue site lists and let and improve targeting? Are all of these providers Network-you determine which sites to avoid? Does it use a verification tool like Advertising-Initiative compliant? How’s the back office? Are theAdSafe or DoubleVerify? If the network uses its own tools,how does it billing systems in order? Will the network work quickly to eliminateprove appropriate placements? A network like Undertone will give a discrepancies and reconcile billing with delivery? What are the net-money-back guarantee (though CEO Michael Cassidy reports the work’s client services like? Despite the advent of Wall Street-like trad-company has never had anyone ask for one). Does the network have ing systems, advertising remains a people business and many agen-some vertical specialty? Contextual content in areas like health can be cies choose ad networks based on trust and sales-rep relationships.valuable, even though placements aren’t on top sites. Buying a verti- Do campaigns deliveragainst goals? If the agency does a pre-postcal network can help lower CPMs and still provide relevant context. buy test, how high was the accuracy in delivering on key perform- Can the network offer scale of rich-media opportunities? Rich ance initiatives? According to John Montgomery, the old direct-media is great,but often a marketer doesn’t just want to make a splash response adage “Test and Learn” applies here. If a network doesn’ton only one or two sites. What is a network’s technological capabili- deliver, drop it.14 | October 11, 2010 |
  15. 15. Sponsored byMEASUREMENT CHART 14 Snapshots of worst malware found in July through DATE: August 2010 for ClickFacts clients 8-31-10 (Hit 30 different Ad Tags) ACCOUNT: A Top 20 PublisherCLICKS: BRANDS EMBRACE RESPONSE MALWARE FOUND: PDF Exploit BEHAVIOR: The ad tag makes a request directly to a URL. The“Only 16% of consumers clicked on at least one display ad in March URL returns Javascript code that contains a link to2009, down from 32% two years earlier.” the malware host site, The intention is–“HOW ONLINE ADVERTISING WORKS: WHITHER THE CLICK,” COMSCORE, 2009 to steal identities.“The click should have been DOA.” DATE: 7-27-10; 7-23-10–RICHY GLASSBERG, COO OF MEDHELP.ORG AND ONE OF THE FOUNDERS OF THE IAB ACCOUNT: Fortune 100 Insurance Company“So many agencies plan on clicks. That’s criminal.” MALWARE FOUND: PDF Exploit–JON GIBS, SENIOR VP-ANALYTICS AND INSIGHT, NIELSEN ONLINE BEHAVIOR: Downloads a PDF that launches an executable file that hits the registry. The intention is to steal“The click is a crutch. Agencies are getting it; clients are not.” identities.–MICHAEL CASSIDY, CEO, UNDERTONE NETWORKS DATE: 7-13-10“64% of advertisers are measuring with clicks.” ACCOUNT: Fortune 100 Car Company–COLLECTIVE MEDIA, 2010 DISPLAY ADVERTISING STUDY MALWARE FOUND: PDF Exploit BEHAVIOR: Downloads a PDF that launches a malicious“The overall click-through rate for static ads is 0.1%, for Flash ads, .09%, process. The intention is to steal identities.consistent with 2008.” – DOUBLECLICK, 2009 YEAR IN REVIEW BENCHMARKS, RELEASED JUNE 2010“If you get the consumer to the point of engagement, it’s a lost DATE: Weeklyopportunity if you don’t ask for the interaction.” ACCOUNT: Seen across multiple accounts–STACEY DEZIEL, MANAGING PARTNER-ACCOUNT DIRECTOR, MEC MALWARE FOUND: Registry updates for Flash & Quicktime BEHAVIOR: A site hits the registry and then asks to upgrade To measure by click,or not to measure by click? It’s a question that Flash and/or Quicktime. While not alwaysthe industry has lived and nearly died by.There’s nothing wrong with malicious, this is a potential vector for infection.soliciting an action to media, but it’s not a sufficient way to pay con- Source: ClickFactstent providers. Research from comScore found that few people areclickers and the percentage of those who do click is in decline. Just16% of online users regularly click on ads,according to a 2009 report. Perhaps the most disturbing finding of a Collective Media study of BEST PRACTICE: BRANDING IMPACT STUDIESmarketers was that the majority were measuring a campaign’s results In early 2010, the IAB and Bain released a study that diagnosed theby clicks.A paltry 35% said they used some sort of brand measure. problem of why brand marketers were not focusing on interactive as Can a click ever be a useful measure for a brand marketer? Stacey much as the IAB thought they should. A group of 700 marketersDeziel of MEC thinks it can be one factor in deciding the success of a were asked which metrics they wanted for online campaigns andcampaign. Clicks as an action within a rich-media engagement, she which metrics they were able to get (see chart 15, P. 17). Somethingbelieves, have value because the person who clicks has demonstrated very strange is going on. The top metrics marketers seek (messagetrue intent. recall, ad favorability and purchase intent) are the very metrics that Despite a terrible reputation, clicks are not going away, and prob- have been around since Nick Nyhan founded Dynamic Logic inably never will in a medium with so much inventory and so many 1999, and are now being cooked into the majority of larger brandingpublishers willing to sell it that way. Many are also using “blended buys. Are marketers just not getting it? The more sophisticated in theCPMs,” buys based on a cost-per-thousand rate so that the ads can field, those such as data modelers Michele Madansky and Yaakovachieve desired reach but with a cost-per-click (CPC) rate for guar- Kimmelfeld, have moved beyond that set of metrics. According toanteed delivery of interested consumers. For benchmarks on click Kimmelfeld:“Once you do a few, you typically know what the resultsperformance by type of ad, including impact of sizes and rich media, are going to be.”see DoubleClick’s recently released “2009 Year in Review But for marketers relatively new to interactive, that data reassuresBenchmarks.” While direct response-focused, the report does make that the medium works.“It’s in the agency’s best interest to do them tothe point that people are interacting with ads,and that larger ads and justify reallocation towards digital,” said Lynn Bolger, exec VP-ad solu-rich media tend to elicit higher click rates. tions at comScore. “They are also helpful when there is no direct con- | October 11, 2010 | 15
  16. 16. BUILDING BRANDS ONLINEnection to offline sales,as in the case of consumer package goods [CPG], “interactive” are ever to break down, studies will need to look atwhere through-panel-matches online data can be connected to offline.” results across all media types. Dynamic Logic now performs cross- Clearly,demand for this type of survey has grown because there are media studies that can show relative contribution of the big three (TV,so many companies in the space. Dynamic Logic pioneered the tech- print and online), as well as incorporate mobile, gaming and socialnique of pop-up surveys to an exposed versus a control group to deter- media to standard brand-impact measures. Dynamic Logic has con-mine the lift generated by the campaign. Dynamic Logic is now part of ducted over 350 of these over the past 10 years (see chart 16, P 17). .TNS Kantar Group and has a normative database so that marketers can Bill Havlena, PHD, research analytics, said clients are using themcompare their performance to those within specific ad categories and as a complement to media-mix modeling. This type of study worksnorms over time.Dynamic Logic has carried out more than 5,700 stud- well when no sufficient historical data on certain media are availableies and tested 190,000 creative campaigns over the past 10 years.Insight or there is relatively low reach for some media on the plan comparedExpress emerged in 1999 as the first alternative. It has conducted more to others. Consumer package goods companies are the biggest usersthan 1,500 studies of 50,000 creative campaigns. Nielsen and comScore of this type of research, but last year saw a big uptick in usage byalso perform the services now. Nielsen has two versions: one for use on financial-services companies. Studies like this cost anywhere fromdisplay advertising,and another for determining whether an ad is elicit- $60,000 to $250,000,depending upon whether Dynamic Logic buildsing conversations on Facebook.Vizu,a Silicon Valley startup that initial- out a simulator for optimizing media mix based on focused on online polling and social network activity, now offers a Havlena’s key insight? “TV, online and print together are veryproduct called Ad Catalyst.Crowd Science,a two-year-old research com- strong at building top-of-mind, unaided brand awareness and com-pany run by former executives from comScore and Nielsen, does pre- municating key brand messages (see chart 17, P. 18).” When used inand post-surveys within their Campaign Audience Profiling Tool. The combination, these media show three times the increase of TV alone.newest entry into the market is Dimestore,from Knowledge Networks, Rex Briggs—who, along with the IAB and the ARF, broughtwhich specializes in the placement of campaign-measurement surveys XMOS, or cross-media optimization studies, to the online world between 2001 and 2006—built a consultancyWhat do marketers need to know?‘It all comes back to based on this type of study, and counts among his clients Honda,Acura, MTV Networks andyou get what you pay for Be wary of the cheap solution .’ the agency RPA. He now has enough datathat yields quick and positive data but little true insight. from his Return on Marketing Objectives Studies (ROMO) that he has built out a nor- mative database and a dashboard tool that willin online video streams, either as overlays or as part of the pre- or post- optimize media spending not on gross ratings points (GRPs) but onroll. Their surveys do not interrupt the video experience or require impact. Eight categories of data are represented, including CPG,respondents to go to a third-party site, and the results are available in auto and financial services. Perhaps the most compelling aspect ofnear real time. the tool is that it integrates and translates GRP data as supplied by What does a branding-impact survey cost? Drew Lipner, exec VP- Telmar, one of the leaders in the television-buy-optimization director of Insight Express, reports that “pricing for ad-effective- The tool, called Matterhorn, launched in beta in late September.ness research is typically categorized by service level, and level of RPA has been using it for the past six rigor.” For do-it-yourself offerings, which areoften a single question, pricing ranges from $5,000 to $9,000. Full-serv- BEST PRACTICE: MEASURE IMPACT OVER TIME AND DETERMINEice research typically includes: project management, survey WHAT IS THE OPTIMAL ONLINE CONVERSION PATHdesign/build, dedicated analyst/project management, media/publisher The challenge with brand-impact studies that include some sort ofcoordination, data weighting, written report and presentation. Pricing “intent” metric is that what people say they will do does not exact-within the full-service segment ranges from $20,000 to $25,000. ly correlate to their actual activity. But interactive excels at provid-According to Lipner, agencies tend to employ full-service brand meas- ing data on what people do online after exposure to an ad—urement, given the need for project management and third-party whether or not they click.Action over time post-exposure withoutaccreditation, along with the desire to use the data to optimize creative. a click is called “view through” and has been a standard online Brand-impact surveys are such a crowded space that they have metric available through third-party ad servers for over a decade.incited the most recent IAB research initiative, which focuses on DoubleClick conducted two waves of research in the mid-’00s thatbest practices in developing the control groups. The IAB press showed which percentage of these “view-throughs” could berelease generated a storm of inside-baseball discussion in the directly attributed to the online exposure; these studies showedonline-research community. What do marketers need to know? As attribution figures around 65%. It’s all a matter of setting up theone poster on the Research Wonks listserve noted:“It all comes back campaign appropriately and setting an appropriate window toto, you get what you pay for.” Be wary of the cheap solution that track those conversions. Microsoft research over the past two yearsyields quick and positive data but little true insight. has shown that too often, the window of tracking a view-through was too short: one week or less. Research from comScore found thatBEST PRACTICE: GO A STEP FURTHER AND MEASURE BRANDING IMPACT ACROSS MEDIA activity continues over at least one month post-exposure, dependingMedia do not exist in isolation, and if the silos of “traditional” and upon the product and the campaign (see chart 18, P. 20).16 | October 11, 2010 |
  17. 17. Sponsored byCHART 15 700 marketers were asked which metrics they wanted for online campaigns, and which metrics they were able to get. PERCENT OF RESPONDENTS 60 WHAT BRAND MARKETERS WANT WHAT BRAND MARKETERS GET #3 #2 30 #1 0 Brand Likelihood to Conversion Click Message Time spent View awareness recommend rates through association on page through Purchase Favorability Recall Unique Ad Interaction Engagement intent visitors impressions/ rate time views Source: Bain/IAB 2009 Marketer SurveyCHART 16 All brand-impact studies operate on similar measurement principles. AWARENESS Brand awareness (measures the level of familiarity respondents have with the brand (aided and unaided) ONLINE AD AWARENESS Measures whether or not respondents recall an advertisement for a brand, product or service Awareness MESSAGE ASSOCIATION Measures the extent to which respondents can match the messages and/or concepts in the creative to the brand BRAND FAVORABILITY Measures the extent to which respondents have a positive or favorable opinion of the brand PURCHASE INTENT Measures the likelihood of Persuasion respondents to purchase the brand in the future Source: Dynamic Logic | October 11, 2010 | 17